Eight RCTs (13,614 participants) were included. Study size ranged from 55 to 9,057 participants. Two studies were cluster RCTs (818 participants) where hospitals were the unit of randomisation and participants joined on hospitalisation and left on discharge, others were parallel studies. Four trials evaluated secondary prevention, three evaluated primary prevention and one evaluated mixed population. Date of publication ranged from 1968 to 1992. Follow-up ranged from an average of one year to eight years. Funnel plot showed potential for publication bias and Begg's test was borderline. Two studies (9,903 participants) scored 3 for quality and six (3,711 participants) scored 2.
Compared to control, risk of coronary heart disease events was reduced with polyunsaturated fatty acid diet (RR 0.81, 95% CI 0.70 to 0.95, I2=37%).
In subgroup analyses, the result was similar to the main analysis for longer term studies, secondary prevention and studies that scored 2 for quality.
Risk of coronary heart disease mortality was reduced with polyunsaturated fatty acid diet (RR 0.80, 95% CI 0.65 to 0.98). There was no effect on total mortality.
In sensitivity analyses, coronary heart disease events were reduced with polyunsaturated fatty acid diet in analyses that excluded studies with open enrolment and those studies that assessed dietary advice. Restriction to double blind trials and to studies that provided meals showed no statistically difference between polyunsaturated fatty acid and control in risk of coronary heart disease events.